Welcome to A Watering Hole Attack,
brought to you by IBM. In this video, you will investigate
the anatomy of a watering hole attack. We can study a real-world attack scenario
to explain the following details. How to instigate a successful
attack by infecting portable computers outside of
an organization's physical network infrastructure using
a watering hole attack? How this infected computer then
spreads the malicious code? And how it contacts a remote command, and control server once it returns to
the organization's environment? How the overall timeline works for the bad
guys, and that this type of attack can only be mitigated by correlation and
collaboration? Or security intelligence inside an
organization using a variety of detection tools across several IT disciplines. This example of a watering hole
attack that took place in 2012 and was subsequently analyzed by
the IBM X-Force research team. Fraudulent malware download maybe
as part of a JPEG, a PDF or just by visiting a website that
downloads a malicious JavaScript that it is not detected
by antivirus software. Next spearfishing, it lure's in people
to click on something interesting. And finally, the network attack vectors,
command and control malware use unusual ports on the clients machine to
communicate with remote control servers. The next slides look at the timeline,
the actual vulnerabilities that were involved in the malicious
communication scheme occurred. In contrast to the length
of the breach for a target, you can see from the timeline
how fast and efficiently the attackers used a zero-day vulnerability to
infiltrate many organizations. In July 13 through 15, several regional consumer financial
services websites were hacked. The hackers plant a hidden
iframe on the consumer portal. During this time frame,
customers of the bank are redirected to a malicious download site when they
visited to do their online banking. Finally infections were
detected at several companies. Let's look at what the vulnerable
hosts were that were infected. Attackers used different
variants of the Gh0st RAT, remote access Trojan horse
making detection very hard. What is Gh0st RAT? Gh0st RAT or remote access terminal is a Trojan remote
access tool used on Windows platform and has been used to hack into some of the
most sensitive computer networks on Earth. Let's take a brief look at
the capabilities in reach of a Gh0st RAT. The Gh0st RAT can take full control of
the remote screen on the infected bot, provide real-time as well as
offline keystroke logging. Can provide live feed of webcam,
microphone of an infected host. Can download remote binaries on
the infected remote host and it can take control of remote shutdown,
or reboot of a host. It can also disable infected computer
remote pointers and keyboard input. Can enter into a shell of remote
infected host with full control. It can provide a list of all
the active processes and clear all existing SSDT
of all existing hooks. So, let's talk about a little bit about
the variance that we saw within this watering hole attack. Variant A exploited a known
Microsoft vulnerability, and was not recognized by
any antivirus vendors. Variant B exploited a known
Java vulnerability. And some patches were available but
not with all antivirus softwares. In this instance after being infected,
compromised hosts may contact with a remote command and
control server in China. Infected machines attempt to communicate
with one of two Chinese command and control, C&C servers. If communications are successfully
established, the C&C server gains complete real-time control of a system
on the protected network. The malware, a remote access Trojan,
allows a remote attacker to access data, log system activity,
capture key logs, take screenshots. Activate the system's camera and
record from the system's microphone. Finally, the remote attacker can
also drop additional downloads and programs in the controlled machine. This is all done for further attacks. If the attack does not
detect it fast enough, the infected machine becomes the new
launch point of deepening the penetration. The following details describe how each
of these attack vectors can be countered by proper measures. The infected machine legitimately
distributes more malware inside the enterprise network to gain
a stronger foothold if detected. Endpoint management negation. Additional software gets installed on
the machine by remote malware and control. Endpoint management software should
immediately detect any new software deployment. Report them, and either remove them or
deny the network access. The malware's first goal is to obtain
privileged user identities which then uses to gain access to valuable
resources inside the enterprise network. Privileged user access. If a machine of a privileged user
is found that their credential is going to open many doors for
the attackers. Control, a privileged user access control
system can negate the chance of any attack or gaining privileged access. Because those IDs have to be signed
out through a particular process using multi-factor authentication and
other security means. Another control could be if privileged
user accesses maliciously gained, a data access monitoring
solution can realize that large amounts of privilege
data is being accessed. And a behavioral pattern that does
not reflect the usual routines and report on it. Most attacks use ports and scans that typically are not executed from
either the infected machines or user IDs. We can look at network anomaly,
unusual ports and scan activity is detected from IT systems
that usually do not display such activity. Another control we can look at is the flow
control system shows traffic records involving on-site and off-site IT systems
and immediately logs and reports this. These attacks are rarely
an isolated event. And the attacked organization is one
of out of many who are being probed by those remote command and control systems. So one control would be public threat
research feeds the recognized IP addresses and ports into a black list of malicious
host that can be incorporated into the organization security
intelligence solution. Only the correlation to
all these events in almost real-time enables
an organization to detect and hopefully stop threats before they can
be exploited and cause any damage. Next we will look at a summary of
those challenges in a broader term. Generally, security intelligence
has focused on real-time or near real-time security analysis. But now new motivations exist for extending the role of
security intelligence. First, data is available to be processed,
security data will need to be persisted for longer times to
detect longer running attack patterns. New cyber data sources have more
security relevance now, such as DNS. Business application data can be
correlated with security data and unstructured content. Second, there is a need for
more advanced analytics. It does not make sense to employ
in a real-time environment. Depth of analysis performed by
sophisticated algorithms such as regression analysis or predictive
algorithms will be longer running and might offer greater security insights. Newer analytic behaviors on the part of
security analysts need to be supported. In the next video, Adam will review
an overview of fishing scams, and I will be back later to
describe the next case study.